To: Michael W.
From: Geoffrey Klempner
Subject: Criteria for the identity of a person over time
Date: 21 March 2002 14:10

Dear Mike,

Thank you for your letter of 3 March with your third essay for Philosophy of Mind, in response to the question, 'Can one give adequate criteria for the identity of a person over time? Illustrate your answer with thought experiments describing "problem cases" of personal identity.'

'Each time we try to establish a baseline upon which to measure change we find that that baseline needs another baseline of its own - ad infinitum!'

I am impressed by the way you have conducted your own original investigation of personal identity, rather than go to the books and repeat the usual things about the 'memory criterion' or 'body criterion' etc.

The way you have laid out the options, both for what you call 'internal identity' and 'external identity' - under the headings 'addition, subtraction division and multiplication' - makes it possible to approach the problem at a considerable degree of abstraction but at the same time very simply.

I would have liked to see more argument for your conclusion, which you state in the lines I have quoted above. It is really just a matter of spelling things out, pointing to the way the 'baseline' shifts in each case.

It is really remarkable, that things to do with what is 'inner', the mind, and things to do with what is 'outer', the body, come under the very same logical categories. How is that possible? Is it just a happy accident, or does it show something about the concept, not of 'inner' or 'outer' but of temporal existence as such?

Looking more closely at the categories, however, I detect possible signs of strain. I think you have tried too hard to relate the categories to examples in the actual world.

Addition and subtraction are easy enough. An entity which exists over time, be it mental or physical, can have bits added on to it or have bits taken away. It is not quite so clear from your examples that 'division' means the same thing in the mental and physical case. Here we need the help of a bit of science fiction.

Mental division, in terms of the problem you are investigating, ought to be something that happens to an entity which was previously undivided. In the mental case, this would happen if all your mental states divided into two packages, each gaining their own identity in the process. The criterion for such division can be stated as follows: A and B are in the same mental packet if and only if every mental item that is present to A is also present to B. If the packet now splits, with A in one packet and B in another, then the same condition can be stated with A and C in one packet and with B and D in another packet. This may or may not be the way some schizophrenic states actually arise: I suspect not.

In the physical case, we would be dealing with human fission, a human being splitting like an amoeba, rather than Siamese twins. Siamese twins do not pose the same challenge for personal identity, because it is generally the case that the twins, though physically connected, are best described as two individuals who share common parts. Where things would get tricky, is if the brain itself was one of the parts/organs that needed to be divided.

Mental multiplication, on the other hand, is something that could be happening right now, although we can never know. As I type these words, a new 'self' comes into existence, thinking exactly the same thoughts and feeling exactly the same feelings, then another then another, until there are thousands of selves each thinking *I* am the one who makes the decision to type the next word. This is actually a powerful argument against a purely inner definition of personal identity: namely, the fact that there is no criterion for *counting* inner 'selves'.

The external case of multiplication is the spectacular example of the person-duplicating machine.

Is that all? Actually, I think not. It could be argued that you have missed out one of the most difficult challenges for the definition of identity, namely, gradual replacement.

Take something neutral, such as the car I used to own. I can add to my Ford Capri, or take things away from it. If Martians point their ray guns at it, it can undergo fission or duplication. But there is also something common or garden which happens to things all the time - and people too. You replace one wheel, then another, and another. New clutch, new water pump, new differential, rewelding here and here and here. Each time we have the 'same' Capri with modifications. Yet it is conceivable that the outcome of this process could be a car which might have no physical bits belonging to the original car.

(Then someone goes to a scrap heap and makes a car out of all the original bits.)

In your terms, gradual replacement is simply the result of successive subtractions and additions. However, these two operations can be combined in a way that poses a special challenge to the idea of identity.

All the best,

Geoffrey